A visual inspection method based on workpiece inspection equipment

Abstract

The present invention discloses a visual inspection method based on workpiece inspection equipment, which relates to the technical field of workpiece inspection. The key points of its technical solution include the following steps: collecting images of a rubber workpiece to be inspected at different visual angles to form a target inspection image set, screening images with defect features in the rubber workpiece from the target inspection image set and marking them as inspection defect images; identifying the defect continuity state, defect category and defect position of the rubber workpiece defect features appearing in the inspection defect image, judging whether it is necessary to perform splicing and synthesis processing on at least two inspection defect images according to the defect category and defect continuity state, and marking the corresponding inspection defect image as an image to be inspected and processed; collecting defect status information of the rubber workpiece in the image to be inspected and processed; the effect is that the present application can inspect the rubber workpiece from multiple angles and is helpful to control product quality.

Description

Technical Field

[0001] The present invention relates to the technical field of workpiece detection, and more particularly, to a visual detection method based on workpiece detection equipment. Background Art

[0002] Rubber workpieces are widely used in numerous fields, such as automotive, machinery, and aerospace. Various defects are inevitable during the production process, and their quality directly impacts the performance, safety, and reliability of related products. Therefore, inspection of rubber workpieces is crucial. With the advancement of computer vision technology, image recognition-based visual inspection has begun to be applied to rubber workpiece inspection. Visual inspection identifies defects by acquiring images and performing simple feature extraction. This method can detect some obvious cosmetic defects, but its detection capabilities are limited for complex defects, such as those with continuous defects or multiple defects interfering with each other. For example, it is difficult to accurately identify rubber molded products with irregular surfaces and multiple overlapping defects. Furthermore, the ability to integrate and process the same defect across multiple images is insufficient, making it difficult to accurately determine the interplay between defects and their combined impact on workpiece performance. Summary of the Invention

[0003] In view of the shortcomings of the prior art, the object of the present invention is to provide a visual inspection method based on workpiece inspection equipment.

[0004] To achieve the above object, the present invention provides the following technical solutions:

[0005] A visual inspection method based on a workpiece inspection device, the method comprising the following steps:

[0006] Collect images of the rubber workpiece to be inspected from different visual angles to form a target detection image set, and select images with defect features in the rubber workpiece from the target detection image set and mark them as detection defect images;

[0007] Identify the defect continuity state, defect category, and defect location of the rubber workpiece defect features in the defect detection image, determine whether it is necessary to perform splicing and synthesis processing on at least two defect detection images based on the defect category and defect continuity state, and mark the corresponding defect detection image as an image to be detected and processed;

[0008] Collecting defect information of the rubber workpiece in the image to be inspected and processed;

[0009] Judging whether there are defect features that affect each other in the rubber workpiece in the image to be inspected based on the defect position and the number of defect categories;

[0010] If there are defect features in the rubber workpiece in the image to be inspected that do not affect each other, a first influence of the defect features on the rubber workpiece in the image to be inspected is determined based on the defect category and defect condition information; if there are at least two defect features in the rubber workpiece in the image to be inspected that affect each other, a second influence of the defect features in the image to be inspected that affect each other is determined based on the defect category and defect condition information;

[0011] Whether the defect characteristic affects the performance of the rubber workpiece is determined according to the first impact condition and / or the second impact condition.

[0012] Preferably, judging whether it is necessary to perform splicing and synthesis processing on at least two defect detection images according to the defect category and the defect continuity state, and marking the corresponding defect detection images as images to be detected and processed, specifically includes the following steps:

[0013] If the continuous defect states of the same defect category appear on the same detection defect image, the detection defect image is marked as an image to be detected and processed;

[0014] If the defect state of the same defect category appears continuously on at least two defect detection images, the corresponding defect detection images are segmented to form a segmented image area to be detected that includes the defect category;

[0015] According to the defect category, defect features at the edges of each segmented image area to be detected are identified, and according to the defect features, the defect edges of each segmented image area to be detected are processed and spliced ​​together to form a processed image to be detected.

[0016] Preferably, according to the defect characteristics, defect edges of each segmented image area to be inspected are processed and then spliced ​​together to form a processed image to be inspected, which specifically includes the following steps:

[0017] Determine whether there is an overlapping portion of defect features at the edges of at least two segmented image regions to be inspected;

[0018] If there is an overlap between the defect features at the edges of at least two segmented image regions to be detected, the corresponding segmented image regions to be detected are marked as matching image regions to be detected, and the corresponding matching image regions to be detected are spliced ​​according to the overlapping defect features to form a processed image to be detected, and the overlapping defect features of one of the matching image regions to be detected are retained, and the overlapping defect features of the remaining matching image regions to be detected are deleted;

[0019] If there is no overlapping part of the defect features at the edge of each segmented image area to be detected, the defect features at the edge of each segmented image area to be detected are paired; if the defect features at the edge of the segmented image area to be detected can be paired so that the defect features at the edge of the corresponding segmented image area to be detected appear to be continuous, the corresponding segmented image area to be detected is spliced ​​to form the processed image to be detected.

[0020] Preferably, judging whether there are defect features that affect each other in the rubber workpiece in the image to be inspected based on the defect positions and the number of defect categories specifically includes the following steps:

[0021] If the rubber workpiece in the image to be inspected has only one defect feature of a defect category, it is determined that the rubber workpiece in the image to be inspected has defect features that do not affect each other;

[0022] If the rubber workpiece in the image to be inspected has defect features of at least two defect categories, determining whether the defect features of the rubber workpiece in the image to be inspected overlap or intersect based on the defect positions;

[0023] If the defect features of the rubber workpiece in the image to be inspected do not overlap or intersect, it is determined that the defect features of the rubber workpiece in the image to be inspected do not affect each other.

[0024] If the defect features of the rubber workpiece in the image to be inspected and processed overlap and intersect, it is determined that the rubber workpiece in the image to be inspected and processed has at least two defect features that affect each other.

[0025] Preferably, determining the first influence of the defect feature on the rubber workpiece in the image to be inspected based on the defect category and defect condition information specifically includes the following steps:

[0026] Comparing the defect condition information of each defect category in the rubber workpiece with the category defect condition-performance table to obtain the first performance impact type and first performance impact degree of each defect category on the rubber workpiece;

[0027] Classifying the first performance impact category of each rubber workpiece into a performance improvement impact category and a first performance reduction impact category;

[0028] The first performance impact type and the first performance impact degree of the rubber workpiece are processed to obtain a first impact situation according to whether the performance improvement impact type and the first performance reduction impact type are the same.

[0029] Preferably, the first performance impact type and the first performance impact degree of the rubber workpiece are processed to obtain the first impact situation according to whether the performance improvement impact type and the first performance reduction impact type are the same, which specifically includes the following steps:

[0030] determining whether there is a same type of performance-enhancing impact category and a first performance-degrading impact category;

[0031] If there are no identical performance-enhancing impact categories and first performance-degrading impact categories, combining the first performance impact categories and corresponding first performance impact degrees of each rubber workpiece to form a first impact situation;

[0032] If there are the same type of performance improvement impact type and first performance reduction impact type, the first performance impact degree of the performance improvement impact type and the first performance impact degree of the first performance reduction impact type are calculated to obtain the first performance impact degree of the same first performance impact type, and the first performance impact degree of the same first performance impact type and the first performance impact degrees of the remaining first performance impact types are combined to form a first impact situation.

[0033] Preferably, judging the second influence of the mutually influencing defect features on the rubber workpiece in the image to be inspected based on the defect category and defect condition information specifically includes the following steps:

[0034] Determine the overlapping and intersecting conditions of the defect features of the rubber workpiece in the image to be inspected and processed;

[0035] If there are superimposed defect features on the rubber workpiece in the image to be inspected, analyzing the performance impact of the rubber workpiece itself based on the superimposed defect features to obtain a first performance impact situation;

[0036] If there are intersecting defect features in the rubber workpiece in the image to be inspected, the performance impact of the rubber workpiece itself is analyzed according to the intersecting defect features to obtain a second performance impact situation;

[0037] The first performance impact situation and / or the second performance impact situation are combined to form a second impact situation.

[0038] Preferably, the performance impact of the rubber workpiece itself is analyzed based on the superimposed defect characteristics to obtain a first performance impact, specifically:

[0039] Determine whether the superimposed defect features cause defects to the rubber workpiece itself. If the superimposed defect features cause defects to the rubber workpiece itself, obtain defect feature type information and defect feature status information, compare the defect feature type information and defect feature status information with the defect type status-performance table to obtain the second performance reduction impact type and second performance reduction impact degree of the rubber workpiece, and mark the second performance reduction impact type and second performance reduction impact degree as the first performance impact situation.

[0040] Preferably, the second performance impact is obtained by analyzing the performance impact of the rubber workpiece itself based on the cross defect characteristics, specifically:

[0041] Compare the defect category and defect condition information to the category defect condition-performance table to obtain the second performance impact category and second performance impact degree of each defect category on the rubber workpiece, and mark the second performance impact category and second performance impact degree as the second performance impact situation.

[0042] An electronic device comprises a memory, a processor and a computer program stored in the memory and runnable on the processor. When the processor executes the program, a visual inspection method based on a workpiece inspection device is implemented.

[0043] Compared with the prior art, the present invention has the following beneficial effects:

[0044] By capturing images of the rubber workpiece to be inspected from different visual angles, this application can obtain comprehensive surface information on the workpiece, thereby avoiding detection blind spots caused by a single viewing angle. It also identifies defect continuity, defect category, and defect location. Based on the defect category and defect continuity, it determines whether to stitch and synthesize the detected defect images. When the same defect category appears continuous across multiple images, it integrates the scattered defect information into a complete defect view through segmentation, edge processing, and stitching. It then determines whether the defect features interact based on the defect location and number of defect categories. Detailed defect impact analysis strategies are developed for different scenarios (non-interactive and interactive). For non-interactive defects, the type and degree of impact of each defect category on the workpiece's performance are determined by comparing them with a defect type-defect-status-performance table, allowing for a comprehensive assessment of the overall impact. For interactive defects, separate analyses are performed based on overlapping or overlapping scenarios, resulting in a more accurate assessment of the impact on the rubber workpiece's performance. Therefore, this application enables multi-angle inspection of rubber workpieces, helps control product quality, and makes reasonable disposition decisions for workpieces with different defect levels, such as repair, downgrade, or scrapping. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 A schematic diagram of the steps of a visual inspection method based on a workpiece inspection device is proposed in the present invention;

[0046] Figure 2 A schematic diagram of the steps for obtaining a second impact condition in a visual inspection method based on a workpiece inspection device proposed by the present invention;

[0047] Figure 3 It is a structural diagram of an electronic device provided by an embodiment of the present invention.

[0048] 610 , processor; 620 , communication interface; 630 , memory; 640 , communication bus. DETAILED DESCRIPTION

[0049] Reference Figures 1 to 3 shown.

[0050] The embodiment further illustrates a visual inspection method based on a workpiece inspection device proposed in the present invention.

[0051] A visual inspection method based on a workpiece inspection device, the method comprising the following steps:

[0052] Collect images of the rubber workpiece to be inspected from different visual angles to form a target detection image set, and select images with defect features in the rubber workpiece from the target detection image set and mark them as detection defect images;

[0053] Identify the defect continuity state, defect category, and defect location of the rubber workpiece defect features in the defect detection image, determine whether it is necessary to perform splicing and synthesis processing on at least two defect detection images based on the defect category and defect continuity state, and mark the corresponding defect detection image as an image to be detected and processed;

[0054] Collecting defect information of the rubber workpiece in the image to be inspected and processed;

[0055] Judging whether there are defect features that affect each other in the rubber workpiece in the image to be inspected based on the defect position and the number of defect categories;

[0056] If there are defect features in the rubber workpiece in the image to be inspected that do not affect each other, a first influence of the defect features on the rubber workpiece in the image to be inspected is determined based on the defect category and defect condition information; if there are at least two defect features in the rubber workpiece in the image to be inspected that affect each other, a second influence of the defect features in the image to be inspected that affect each other is determined based on the defect category and defect condition information;

[0057] Whether the defect characteristic affects the performance of the rubber workpiece is determined according to the first impact condition and / or the second impact condition.

[0058] This application utilizes the workpiece inspection equipment's visual system to photograph rubber workpieces from multiple angles. Because rubber workpieces may have defects of varying shapes and locations, a single viewing angle cannot fully capture them. Shooting from different angles ensures that every surface of the workpiece is captured, thereby forming a target detection image set. Within the target detection image set, each image is screened for defect features based on defect feature recognition rules (such as color anomalies, shape deviations, and texture changes), and these images are marked as detected defect images.

[0059] Image processing and recognition technology is used to identify the rubber workpiece defect characteristics, including the defect continuity state (e.g., whether the defect is linearly continuous or discretely distributed), defect type (e.g., different types of defects such as cracks, bubbles, wear, and bumps), and defect location (specifically, where on the workpiece). Based on the defect type and continuity state, the system determines whether the defect images need to be spliced ​​and synthesized.

[0060] If the continuous state of defects of the same defect category appears on the same defect detection image, it means that the image already contains the relevant information of the defect, and it will be marked as the image to be detected and processed. When the continuous state of defects of the same defect category appears on at least two defect detection images, since the defect information is scattered in multiple images, these images need to be segmented to determine the segmented image area to be detected containing the defect category. The defect features at the edges of each segmented image area to be detected are identified through technical means such as edge detection. In order to achieve seamless splicing of images, the defect edges need to be processed (for example, adjusting the grayscale and contrast of the edges to make the transition natural). Then, the processed segmented image areas to be detected are spliced ​​and synthesized to form the image to be detected and processed. In this way, the scattered defect information is integrated into one image, which is convenient for subsequent processing and analysis.

[0061] Defect information is collected from the rubber workpiece in the image being inspected, including but not limited to defect size, shape, and color depth. This information is crucial for accurately determining the impact of defects on workpiece performance. Defect locations and the number of defect types are then used to determine whether the defect characteristics are interdependent.

[0062] If the rubber workpiece in the image to be inspected and processed only contains defect features of one defect category, and there is clearly no mutual interference between these defects, then it is determined that there are independent defect features. If there are defect features of at least two defect categories, the location of the defects is further determined to determine whether there is overlap or intersection. If the defect features do not overlap or intersect in space, they are relatively independent and are also determined to be independent defect features. If overlap or intersection is found, then it is determined that there are at least two mutually affecting defect features.

[0063] The defect condition information for each defect category in the rubber workpiece is compared with a pre-established table of defect condition and performance. This performance table, developed through extensive experimentation, testing, and actual production experience, details the impact of different defect categories on the rubber workpiece's performance under various conditions. This comparison determines the primary performance impact type (e.g., whether it improves performance, degrades performance, or has no significant impact) and the primary performance impact degree (a quantified numerical value of the impact degree) for each defect category. The primary performance impact type for each rubber workpiece is then divided into a performance-improving impact type and a first performance-degrading impact type. A determination is then made as to whether the performance-improving impact type and the first performance-degrading impact type are identical. If no identical types exist, the first performance impact type and corresponding first performance impact degree for each rubber workpiece are directly combined to form the primary impact scenario. If identical types exist, the first performance impact degrees of the performance-improving impact type and the first performance-degrading impact type are calculated (e.g., weighted by weight) to obtain the first performance impact degree for the same first performance impact type. This is then combined with the first performance impact degrees of the remaining first performance impact types to form the primary impact scenario. This results in a comprehensive assessment of the impact of non-interfering defect characteristics on the rubber workpiece's performance.

[0064] The overlap of defect features of the rubber workpiece in the image to be inspected is determined. If overlapping defect features exist, further determination is made as to whether these overlapping defect features cause defects in the rubber workpiece itself. If defects exist (including pits, cracks, and pores), information on the defect feature type and state is obtained. This information is then compared with a defect type and state-performance table (created specifically for defect conditions) to determine the second type and degree of performance degradation of the rubber workpiece, which is then labeled as a first performance impact condition. If overlapping defect features exist, the defect category and state information are compared with the defect type and state-performance table to determine the second type and degree of performance degradation of each defect category on the rubber workpiece, which is then labeled as a second performance impact condition. Finally, the first performance impact condition and / or the second performance impact condition are combined to form a second impact condition. Finally, a determination is made based on the first impact condition and / or the second impact condition as to whether the defect feature affects the performance of the rubber workpiece, thus completing the entire inspection and analysis process.

[0065] Determine whether to perform splicing and synthesis processing on at least two defect detection images based on the defect category and defect continuity state, and mark the corresponding defect detection image as an image to be detected and processed, specifically including the following steps:

[0066] If the continuous defect states of the same defect category appear on the same detection defect image, the detection defect image is marked as an image to be detected and processed;

[0067] If the defect state of the same defect category appears continuously on at least two defect detection images, the corresponding defect detection images are segmented to form a segmented image area to be detected that includes the defect category;

[0068] According to the defect category, defect features at the edges of each segmented image area to be detected are identified, and according to the defect features, the defect edges of each segmented image area to be detected are processed and spliced ​​together to form a processed image to be detected.

[0069] If the continuous defect states of the same defect category appear on the same detection defect image, it means that the image can fully present the relevant information of this defect, and the detection defect image is directly marked as the image to be detected and processed.

[0070] When defects of the same defect category appear continuously in at least two defect inspection images, the defect information is dispersed across multiple images, making it difficult to fully understand the defect from a single image. In this case, these defect inspection images containing defects are segmented. Segmentation regions are determined based on defect category-related features (such as the shape, color, and texture of the defect in different images), forming a segmented image region containing the defect category.

[0071] For each segmented image region to be inspected, image recognition technology is used to identify defect features at the edges of known defect categories. For example, if the defect is a crack, the direction, endpoints, and other features of the crack are identified at the edges of each segmented region. After identification, the defect edges are processed (including adjusting the grayscale value and contrast of the edges) to minimize the visual difference between adjacent segmented regions. Alternatively, the edges are smoothed to eliminate the discontinuity caused by segmentation and create conditions for image stitching.

[0072] After edge processing, the image segments are stitched together based on the identified defect features. Segmented regions are arranged sequentially according to the defect's continuity and then fused together to form a complete image. For example, if the defect is a continuous linear defect, the segments are stitched together along the defect's direction so that the defect appears continuous and complete in the new image.

[0073] According to the defect characteristics, the defect edges of each segmented image area to be inspected are processed and then spliced ​​together to form a processed image to be inspected, which specifically includes the following steps:

[0074] Determine whether there is an overlapping portion of defect features at the edges of at least two segmented image regions to be inspected;

[0075] If there is an overlap between the defect features at the edges of at least two segmented image regions to be detected, the corresponding segmented image regions to be detected are marked as matching image regions to be detected, and the corresponding matching image regions to be detected are spliced ​​according to the overlapping defect features to form a processed image to be detected, and the overlapping defect features of one of the matching image regions to be detected are retained, and the overlapping defect features of the remaining matching image regions to be detected are deleted;

[0076] If there is no overlapping part of the defect features at the edge of each segmented image area to be detected, the defect features at the edge of each segmented image area to be detected are paired; if the defect features at the edge of the segmented image area to be detected can be paired so that the defect features at the edge of the corresponding segmented image area to be detected appear to be continuous, the corresponding segmented image area to be detected is spliced ​​to form the processed image to be detected.

[0077] This application determines whether there is overlap between the defect features at the edges of each segmented image region to be inspected. This step uses image recognition and comparison technology to determine the spatial relationship between defect features at the edges of different regions by comparing factors such as the shape, position, and size of the defect features.

[0078] When defect features at the edges of at least two segmented image regions to be inspected overlap, these overlapping regions are marked as matching image regions to be inspected. Then, stitching is performed based on the defect features in these overlapping regions. To avoid interference caused by duplicate defect features, the overlapping defect features of one matching image region to be inspected are retained, while the overlapping defect features of the remaining matching image regions to be inspected are deleted. This ensures the integrity of defect information while eliminating redundancy, ensuring that the stitched image to be inspected accurately reflects the defect situation.

[0079] If the defect features at the edges of each segmented image region to be inspected do not overlap, these edge defect features are paired. By identifying information such as the shape and direction of the defect features, the defect features at the edges of different regions are combined to form a continuous state. If the pairing is successful, so that the defect features at the edges of the corresponding segmented image regions to be inspected form a continuous defect shape, the segmented image regions to be inspected are spliced ​​together to form the processed image to be inspected.

[0080] The method of determining whether there are defect features that affect each other in the rubber workpiece in the image to be inspected based on the defect locations and the number of defect categories includes the following steps:

[0081] If the rubber workpiece in the image to be inspected has only one defect feature of a defect category, it is determined that the rubber workpiece in the image to be inspected has defect features that do not affect each other;

[0082] If the rubber workpiece in the image to be inspected has defect features of at least two defect categories, determining whether the defect features of the rubber workpiece in the image to be inspected overlap or intersect based on the defect positions;

[0083] If the defect features of the rubber workpiece in the image to be inspected do not overlap or intersect, it is determined that the defect features of the rubber workpiece in the image to be inspected do not affect each other.

[0084] If the defect features of the rubber workpiece in the image to be inspected and processed overlap and intersect, it is determined that the rubber workpiece in the image to be inspected and processed has at least two defect features that affect each other.

[0085] This application first observes the number of defect categories of the rubber workpiece in the image to be inspected. If there is only one defect feature of the defect category, since there is no possibility of interaction between multiple defect categories, it is directly determined that there are defect features that do not affect each other in the rubber workpiece in the image to be inspected.

[0086] When the rubber workpiece contains defect features of at least two defect categories in the image to be inspected, the defect locations are determined. Image recognition and spatial analysis techniques are used to determine the specific location distribution of the different defect features on the rubber workpiece and to determine whether these defect features overlap or intersect. Overlapping refers to the partial or complete overlap of one defect with another; intersecting refers to the spatial interspersal and intersection of different defects.

[0087] If the defect features do not overlap or intersect, it indicates that the different defects are relatively independent in space and do not directly interact with each other. Therefore, it is determined that the rubber workpiece in the image to be inspected has defect features that do not affect each other. However, if the defect features overlap or intersect, it indicates that at least two defects are spatially related and interact with each other. Therefore, it is determined that the rubber workpiece in the image to be inspected has at least two defect features that affect each other.

[0088] Then, the first influence of the defect feature on the rubber workpiece in the image to be inspected is determined according to the defect category and defect condition information, specifically including the following steps:

[0089] Comparing the defect condition information of each defect category in the rubber workpiece with the category defect condition-performance table to obtain the first performance impact type and first performance impact degree of each defect category on the rubber workpiece;

[0090] Classifying the first performance impact category of each rubber workpiece into a performance improvement impact category and a first performance reduction impact category;

[0091] The first performance impact type and the first performance impact degree of the rubber workpiece are processed to obtain a first impact situation according to whether the performance improvement impact type and the first performance reduction impact type are the same.

[0092] This application compares defect condition information (such as defect size, shape, location, and severity) for each defect category in a rubber workpiece with a pre-established table linking defect condition and performance. This table, developed through extensive experimentation, simulations, and actual production experience, details the impact of different defect categories on rubber workpiece performance under various conditions. By comparing each defect category, the primary performance impact (e.g., reduced tensile strength, improved wear resistance) and the degree of primary performance impact (expressed as a specific numerical value or grade) are determined, thereby quantifying the defect's impact on performance.

[0093] The primary performance impact category for each rubber workpiece is categorized as either a performance-enhancing impact category or a first performance-degrading impact category. This is done to differentiate the impacts of defects of different natures, facilitating a more organized analysis of their combined effects on workpiece performance. For example, some defects may increase the elasticity of a rubber workpiece, thus falling into the performance-enhancing impact category; while others may decrease the strength of the workpiece, thus falling into the first performance-degrading impact category.

[0094] The first performance impact type and first performance impact degree of the rubber workpiece are processed based on whether the type of performance improvement and the first type of performance reduction are the same. If the type of performance improvement and the first type of performance reduction do not match, it means that the impact of each defect is relatively independent. The first performance impact type and the corresponding first performance impact degree of each rubber workpiece are combined to form the first impact situation. If the type of performance improvement and the first type of performance reduction do match, it means that there is both improvement and reduction in the same performance aspect. In this case, the first performance impact degree of the performance improvement type and the first performance reduction type need to be calculated (for example, weighted calculation based on certain weights, which can be determined based on factors such as the severity of the defect and the frequency of occurrence) to obtain the first performance impact degree of the same first performance impact type. This is then combined with the first performance impact degrees of the remaining first performance impact types to form the first impact situation.

[0095] According to whether the performance improvement impact type and the first performance reduction impact type are the same, the first performance impact type and the first performance impact degree of the rubber workpiece are processed to obtain a first impact situation, specifically including the following steps:

[0096] determining whether there is a same type of performance-enhancing impact category and a first performance-degrading impact category;

[0097] If there are no identical performance-enhancing impact categories and first performance-degrading impact categories, combining the first performance impact categories and corresponding first performance impact degrees of each rubber workpiece to form a first impact situation;

[0098] If there are the same type of performance improvement impact type and first performance reduction impact type, the first performance impact degree of the performance improvement impact type and the first performance impact degree of the first performance reduction impact type are calculated to obtain the first performance impact degree of the same first performance impact type, and the first performance impact degree of the same first performance impact type and the first performance impact degrees of the remaining first performance impact types are combined to form a first impact situation.

[0099] This application first compares the performance-improving and first performance-degrading categories of the rubber workpiece one by one. By determining the performance-improving category corresponding to each defect category, it is determined whether the same performance-improving and first performance-degrading categories exist. For example, it is determined whether defects that increase and decrease the tensile strength of the rubber workpiece are both present.

[0100] If there are no identical performance-improving impact categories and first performance-degrading impact categories, it indicates that each defect has relatively independent impacts on the rubber workpiece's performance in different aspects. In this case, simply combine the first performance impact categories and corresponding first performance impact degrees for each rubber workpiece. For example, if one defect improves the wear resistance of a rubber workpiece (the first performance impact category is improved wear resistance, and the first performance impact degree is a specific value), while another defect reduces the workpiece's aging resistance (the first performance impact category is reduced aging resistance, and the first performance impact degree is a corresponding value), combining these two impact categories and impact degrees will form the first impact scenario.

[0101] When there are the same type of performance-enhancing impact type and the first performance-reducing impact type, it means that there are both positive and negative impacts on the same performance aspect. In this case, it is necessary to calculate the first performance impact degree of the performance-enhancing impact type and the first performance impact degree of the first performance-reducing impact type. The calculation method can be a weighted calculation. The weights are determined according to factors such as the severity of the defect and the frequency of occurrence. The improvement and reduction impact degrees are comprehensively calculated to obtain the first performance impact degree of the same first performance impact type. This calculated impact degree is then combined with the first performance impact degrees of the remaining first performance impact types to form the first impact situation. For example, for tensile strength, there are both defects that increase and defects that reduce it. After calculating the comprehensive impact degree of the tensile strength performance impact type, it is combined with other performance impact types (such as hardness, elasticity, etc.) to fully reflect the overall effect of the defect on the performance of the rubber workpiece.

[0102] Then, judging the second influence of the defect features that affect each other on the rubber workpiece in the image to be inspected based on the defect category and defect condition information, specifically includes the following steps:

[0103] Determine the overlapping and intersecting conditions of the defect features of the rubber workpiece in the image to be inspected and processed;

[0104] If there are superimposed defect features on the rubber workpiece in the image to be inspected, analyzing the performance impact of the rubber workpiece itself based on the superimposed defect features to obtain a first performance impact situation;

[0105] If there are intersecting defect features in the rubber workpiece in the image to be inspected, the performance impact of the rubber workpiece itself is analyzed according to the intersecting defect features to obtain a second performance impact situation;

[0106] The first performance impact situation and / or the second performance impact situation are combined to form a second impact situation.

[0107] This application first determines the overlap and intersection of the defect features of the rubber workpiece in the image to be inspected and processed, and uses image recognition technology to determine whether different defects overlap (overlap) or intersect (intersect) in spatial position.

[0108] If there are superimposed defect features, the influence of the superimposed defects on the performance of the rubber workpiece is determined, and the performance influence of the rubber workpiece itself is analyzed according to the superimposed defect features to obtain a first performance influence situation.

[0109] When there are intersecting defect features, the path and method of the impact of the intersecting defects on the performance of the rubber workpiece are determined. By judging the interaction of defects at the intersection, such as the possibility that crack intersection may lead to accelerated crack propagation, the second performance impact is obtained based on theoretical knowledge such as material mechanics and rubber processing.

[0110] Combine the primary performance impact and / or secondary performance impact. Based on the actual defect characteristics (primarily the primary performance impact if only overlapping defects exist; primarily the secondary performance impact if only intersecting defects exist; and a combination of both if both exist), comprehensively consider the overall impact of different types of mutually influencing defects on the performance of the rubber workpiece, ultimately forming the secondary impact scenario. This allows for a complete assessment of the performance impact of rubber workpieces with mutually influencing defect characteristics.

[0111] The first performance impact is obtained by analyzing the performance impact of the rubber workpiece itself based on the superimposed defect characteristics, specifically:

[0112] Determine whether the superimposed defect features cause defects to the rubber workpiece itself. If the superimposed defect features cause defects to the rubber workpiece itself, obtain defect feature type information and defect feature status information, compare the defect feature type information and defect feature status information with the defect type status-performance table to obtain the second performance reduction impact type and second performance reduction impact degree of the rubber workpiece, and mark the second performance reduction impact type and second performance reduction impact degree as the first performance impact situation.

[0113] This application utilizes high-precision imaging equipment (such as industrial cameras and microscopes) to capture images of overlapping defect areas, observing at both macro and micro levels to determine whether the overlapping defect characteristics cause abnormalities on the surface or within the rubber workpiece. Macroscopically, the application determines whether there are noticeable irregularities or color differences. Microscopically, the application uses electron microscopy and other methods to determine whether the rubber's molecular structure is damaged or whether there are microcracks. For example, if a rubber seal has overlapping air bubble defects and wear defects, the application determines whether the wear area has more severe surface unevenness due to the presence of air bubbles, and whether microcracks appear around the air bubbles to determine whether a defect has occurred.

[0114] If a defect is detected, the system begins to acquire information on the defect's type and state. This information primarily identifies the defect's type, such as surface pits and dents, internal voids, or delamination. Information on the defect's state includes the defect's size (such as pit diameter or void volume), shape (circular, elliptical, or irregular), location (specific coordinates on the workpiece surface or depth within the workpiece), and quantity. For example, for internal void defects, X-ray tomography is used to determine the specific location and size of the defect's three-dimensional shape within the workpiece.

[0115] The acquired defect type and state information is compared with a pre-established defect type and state-performance table. This performance table is derived from extensive experiments, in which various properties of rubber workpieces (such as tensile strength, elastic modulus, sealing performance, and wear resistance) are tested for defects of different types and states. For example, if the defect characteristics are the volume and number of internal voids, consulting the performance table reveals that this will lead to a decrease in the tensile strength and sealing performance of the rubber workpiece. The magnitude of this decrease corresponds to the specific value of the second performance impact reduction, and the decrease in tensile strength and sealing performance is the second performance impact type. The determined second performance impact type and second performance impact reduction are marked as the first performance impact situation, thus completing the analysis of the impact of the superimposed defect characteristics on the rubber workpiece performance.

[0116] The second performance impact is obtained by analyzing the performance impact of the rubber workpiece itself based on the cross-defect characteristics, specifically:

[0117] Compare the defect category and defect condition information to the category defect condition-performance table to obtain the second performance impact category and second performance impact degree of each defect category on the rubber workpiece, and mark the second performance impact category and second performance impact degree as the second performance impact situation.

[0118] This application compiles information about cross-defect features on rubber workpieces in the image being inspected. Image recognition technology is used to identify defect categories (such as cracks, bubbles, and wear), while also recording detailed information about the defect's size, shape, location, and depth. For example, for cross-defects like cracks and bubbles, the length, width, and direction of the cracks, as well as the diameter, number, and distribution of the bubbles, are measured.

[0119] Compare the organized defect category and condition information to a pre-established table linking defect category and condition to performance. This performance table, derived from extensive experiments and actual production data, details the impact of different defect categories on various rubber workpiece properties (such as tensile strength, elasticity, wear resistance, and aging resistance) under various conditions. This comparison determines the type of secondary performance impact (e.g., reduced tensile strength, improved wear resistance, etc.) and the degree of secondary performance impact (quantified using a specific value or level) for each defect category. For example, comparing the length and depth of a crack defect to the performance table reveals that it reduces the tensile strength of the rubber workpiece. This reduction in tensile strength is identified as a secondary performance impact, and the corresponding secondary performance impact value is obtained.

[0120] The resulting secondary performance impact categories and degrees are integrated and labeled to form a secondary performance impact profile. This result comprehensively reflects the overall impact of cross-defect characteristics on the performance of the rubber workpiece, providing a key basis for subsequently determining whether the defects have a substantial impact on the performance of the rubber workpiece.

[0121] Determine whether the defect characteristics affect the performance of the rubber workpiece based on the first impact situation and / or the second impact situation, specifically:

[0122] When the rubber workpiece in the image to be inspected has only one defect category, or when there are multiple defect categories but they do not have the same type of improved and reduced performance impact (that is, the improved performance impact category and the first reduced performance impact category do not overlap), the judgment is made directly based on the obtained first performance impact category and first performance impact degree.

[0123] If the first performance impact type is "Reduced Performance" and the first performance impact exceeds a pre-defined acceptable threshold (this threshold is determined based on the specific application scenario and quality standards of the rubber workpiece. For example, rubber seals used in aerospace have extremely high performance requirements, so the acceptable performance reduction may be very small; while the threshold for ordinary civilian rubber products is relatively loose), the defect characteristic is determined to have affected the performance of the rubber workpiece. For example, if the impact of the reduction in tensile strength of the rubber workpiece exceeds the specified 5%, it can be considered that the performance is affected.

[0124] If the first performance impact category is "Improved Performance" or "No Significant Impact," and the impact is within a reasonable range (does not cause the workpiece's performance indicators to exceed normal fluctuations), the defect characteristics are considered to have no negative impact on the rubber workpiece's performance. For example, if a defect slightly improves the rubber workpiece's wear resistance while maintaining other performance indicators, performance is considered to be unaffected.

[0125] If there are the same type of performance-enhancing impact category and the same first performance-degrading impact category, the first performance impact degree of the same first performance impact category is combined with the first performance impact degrees of the other first performance impact categories to comprehensively consider the overall performance change.

[0126] Determine the combined performance change resulting from all defect effects and compare it to the performance standard for the rubber workpiece. If the combined performance change causes key performance indicators (such as tensile strength, hardness, and elastic modulus) to deviate significantly from the standard values ​​and exceed acceptable ranges (e.g., tensile strength is less than 90% of the standard value), the defect characteristics are considered to have affected the performance of the rubber workpiece.

[0127] If the overall performance remains within an acceptable fluctuation range, or if the overall performance shows an improvement trend and is within reasonable limits (e.g., each performance indicator fluctuates but still meets the usage requirements), it is determined that there is no substantial negative impact on performance.

[0128] When the defect features of the rubber workpiece in the image to be inspected and processed overlap and produce a flaw, the obtained second performance degradation impact type and second performance degradation impact degree represent the impact of the flaw on the workpiece performance.

[0129] Compare the second performance impact factor to the acceptable standards for the corresponding performance indicators of the rubber workpiece. If the second performance impact factor causes the relevant performance of the workpiece (such as sealing performance and pressure resistance) to fall below the acceptable standards (for example, the leakage rate of a seal due to a defect exceeds the specified minimum leakage rate standard), then the defect characteristic is determined to have affected the performance of the rubber workpiece.

[0130] If the relevant performance is still maintained above the standard, that is, the impact of the defect on the performance is within an acceptable range, then it is determined that there is no impact on the performance; for the case of overlapping defect characteristics, the judgment is made based on the obtained second performance impact type and second performance impact degree.

[0131] If the second performance impact causes an unacceptable decline in the workpiece's key performance (such as dynamic fatigue performance, media resistance, etc.) (such as accelerated aging in a specific medium, exceeding the specified aging rate), it is determined that the defect characteristics have an impact on the performance of the rubber workpiece.

[0132] If the performance change is within an acceptable range (for example, each performance indicator can still meet normal use despite the change), it is determined that there is no impact on performance.

[0133] In actual applications, the first and second impact situations may exist at the same time. The changes in various performance indicators involved in the first and second impact situations are determined, the importance of different performance indicators is considered (for example, for rubber workpieces that bear pressure, the weight of the compressive performance indicator can be set higher), and the comprehensive performance change value is determined.

[0134] Compare the overall performance change to the rubber workpiece's overall performance requirements. If the overall results indicate a performance degradation beyond an acceptable range, affecting the workpiece's normal use and quality standards, the defect is considered to have affected the rubber workpiece's performance. If the overall performance still meets the requirements, it is considered to have no impact on performance.

[0135] An electronic device includes a memory, a processor, and a computer program stored in the memory and runnable on the processor. When the processor executes the program, a visual inspection method based on a workpiece inspection device is implemented.

[0136] The electronic device may include a processor 610, a communication interface 620, a memory 630, and a communication bus 640. The processor 610, the communication interface 620, and the memory 630 communicate with each other via the communication bus 640. The processor 610 may invoke logic instructions in the memory 630 to execute a visual inspection method based on a workpiece inspection device.

[0137] Furthermore, the logic instructions in the aforementioned memory 630 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the portion that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as a USB flash drive, a mobile hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0138] On the other hand, the present invention also provides a computer program product, which includes a computer program. The computer program can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute a visual inspection method based on a workpiece inspection device.

[0139] In yet another aspect, the present invention further provides a non-transitory computer-readable storage medium having a computer program stored thereon, which is configured to execute a visual inspection method based on a workpiece inspection device when the computer program is executed by a processor.

[0140] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, i.e., they may be located in one location or distributed across multiple network units. Some or all of the modules may be selected based on actual needs to achieve the objectives of the present embodiment. Persons of ordinary skill in the art will be able to understand and implement the present invention without inventive effort.

[0141] Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented using software plus a necessary general-purpose hardware platform, or of course, hardware. Based on this understanding, the essence of the above technical solution, or the portion that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a magnetic disk, or an optical disk, and includes a number of instructions for causing a computer device (such as a personal computer, server, or network device) to execute the methods described in each embodiment or certain portions of the embodiments.

[0142] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A visual inspection method based on workpiece inspection equipment, characterized in that: The method comprises the following steps: Collect images of the rubber workpiece to be inspected from different visual angles to form a target detection image set, and select images with defect features in the rubber workpiece from the target detection image set and mark them as detection defect images; Identify the defect continuity state, defect category, and defect location of the rubber workpiece defect features in the defect detection image, determine whether it is necessary to perform splicing and synthesis processing on at least two defect detection images based on the defect category and defect continuity state, and mark the corresponding defect detection image as an image to be detected and processed; Collecting defect information of the rubber workpiece in the image to be inspected and processed; Judging whether there are defect features that affect each other in the rubber workpiece in the image to be inspected based on the defect position and the number of defect categories; If there are defect features in the rubber workpiece in the image to be inspected that do not affect each other, determining a first influence of the defect features on the rubber workpiece in the image to be inspected based on the defect category and defect condition information; Comparing the defect condition information of each defect category in the rubber workpiece with the category defect condition-performance table to obtain the first performance impact type and first performance impact degree of each defect category on the rubber workpiece; Classifying the first performance impact category of each rubber workpiece into a performance improvement impact category and a first performance reduction impact category; According to whether the performance improvement impact type and the first performance reduction impact type are the same, the first performance impact type and the first performance impact degree of the rubber workpiece are processed to obtain a first impact situation; If there are at least two mutually affecting defect features in the rubber workpiece in the image to be inspected, determining a second influence of the mutually affecting defect features on the rubber workpiece in the image to be inspected based on the defect category and defect condition information; Whether the defect characteristic affects the performance of the rubber workpiece is determined according to the first impact condition and / or the second impact condition.

2. A visual inspection method based on a workpiece inspection device according to claim 1, characterized in that: Determine whether to perform splicing and synthesis processing on at least two defect detection images based on the defect category and defect continuity state, and mark the corresponding defect detection image as an image to be detected and processed, specifically including the following steps: If the continuous defect states of the same defect category appear on the same detection defect image, the detection defect image is marked as an image to be detected and processed; If the defect state of the same defect category appears continuously on at least two defect detection images, the corresponding defect detection images are segmented to form a segmented image area to be detected that includes the defect category; According to the defect category, defect features at the edges of each segmented image area to be detected are identified, and according to the defect features, the defect edges of each segmented image area to be detected are processed and spliced ​​together to form a processed image to be detected.

3. A visual inspection method based on a workpiece inspection device according to claim 2, characterized in that: According to the defect characteristics, the defect edges of each segmented image area to be inspected are processed and then spliced ​​together to form a processed image to be inspected, which specifically includes the following steps: Determine whether there is an overlapping portion of defect features at the edges of at least two segmented image regions to be inspected; If there is an overlap between the defect features at the edges of at least two segmented image regions to be detected, the corresponding segmented image regions to be detected are marked as matching image regions to be detected, and the corresponding matching image regions to be detected are spliced ​​according to the overlapping defect features to form a processed image to be detected, and the overlapping defect features of one of the matching image regions to be detected are retained, and the overlapping defect features of the remaining matching image regions to be detected are deleted; If there is no overlapping part of the defect features at the edge of each segmented image area to be detected, the defect features at the edge of each segmented image area to be detected are paired; if the defect features at the edge of the segmented image area to be detected can be paired so that the defect features at the edge of the corresponding segmented image area to be detected appear to be continuous, the corresponding segmented image area to be detected is spliced ​​to form the processed image to be detected.

4. The visual inspection method based on workpiece inspection equipment according to claim 1, characterized in that: The method of determining whether there are defect features that affect each other in the rubber workpiece in the image to be inspected based on the defect locations and the number of defect categories includes the following steps: If the rubber workpiece in the image to be inspected has only one defect feature of a defect category, it is determined that the rubber workpiece in the image to be inspected has defect features that do not affect each other; If the rubber workpiece in the image to be inspected has defect features of at least two defect categories, determining whether the defect features of the rubber workpiece in the image to be inspected overlap or intersect based on the defect positions; If the defect features of the rubber workpiece in the image to be inspected do not overlap or intersect, it is determined that the defect features of the rubber workpiece in the image to be inspected do not affect each other. If the defect features of the rubber workpiece in the image to be inspected and processed overlap and intersect, it is determined that the rubber workpiece in the image to be inspected and processed has at least two defect features that affect each other.

5. A visual inspection method based on workpiece inspection equipment according to claim 4, characterized in that: According to whether the performance improvement impact type and the first performance reduction impact type are the same, the first performance impact type and the first performance impact degree of the rubber workpiece are processed to obtain a first impact situation, specifically including the following steps: determining whether there is a same type of performance-enhancing impact category and a first performance-degrading impact category; If there are no identical performance-enhancing impact categories and first performance-degrading impact categories, combining the first performance impact categories and corresponding first performance impact degrees of each rubber workpiece to form a first impact situation; If there are the same type of performance improvement impact type and first performance reduction impact type, the first performance impact degree of the performance improvement impact type and the first performance impact degree of the first performance reduction impact type are calculated to obtain the first performance impact degree of the same first performance impact type, and the first performance impact degree of the same first performance impact type and the first performance impact degrees of the remaining first performance impact types are combined to form a first impact situation.

6. A visual inspection method based on a workpiece inspection device according to claim 4, characterized in that: Then, judging the second influence of the defect features that affect each other on the rubber workpiece in the image to be inspected based on the defect category and defect condition information, specifically includes the following steps: Determine the overlapping and intersecting conditions of the defect features of the rubber workpiece in the image to be inspected and processed; If there are superimposed defect features on the rubber workpiece in the image to be inspected, analyzing the performance impact of the rubber workpiece itself based on the superimposed defect features to obtain a first performance impact situation; If there are intersecting defect features in the rubber workpiece in the image to be inspected, the performance impact of the rubber workpiece itself is analyzed according to the intersecting defect features to obtain a second performance impact situation; The first performance impact situation and / or the second performance impact situation are combined to form a second impact situation.

7. A visual inspection method based on a workpiece inspection device according to claim 6, characterized in that: The first performance impact is obtained by analyzing the performance impact of the rubber workpiece itself based on the superimposed defect characteristics, specifically: Determine whether the superimposed defect features cause defects to the rubber workpiece itself. If the superimposed defect features cause defects to the rubber workpiece itself, obtain defect feature type information and defect feature status information, compare the defect feature type information and defect feature status information with the defect type status-performance table to obtain the second performance reduction impact type and second performance reduction impact degree of the rubber workpiece, and mark the second performance reduction impact type and second performance reduction impact degree as the first performance impact situation.

8. The visual inspection method based on workpiece inspection equipment according to claim 7, characterized in that: The second performance impact is obtained by analyzing the performance impact of the rubber workpiece itself based on the cross-defect characteristics, specifically: Compare the defect category and defect condition information to the category defect condition-performance table to obtain the second performance impact category and second performance impact degree of each defect category on the rubber workpiece, and mark the second performance impact category and second performance impact degree as the second performance impact situation.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, the visual inspection method based on the workpiece inspection device as described in any one of claims 1 to 8 is implemented.

Images